OS Week Highlights - Oct 2 - 8
-
π173
Open-Orca/Mistral-7B-OpenOrca
Text Generation β’ Updated β’ 4.67k β’ 687Note Mistral model fine-tuned on the OpenOrca dataset
teknium/CollectiveCognition-v1.1-Mistral-7B
Text Generation β’ Updated β’ 25 β’ 77Note Another Mistral fine-tune with great results in TruthfulQA
stabilityai/stablelm-3b-4e1t
Text Generation β’ 3B β’ Updated β’ 16.6k β’ 312Note Very high performant model by Stability. WIth just 3B params, it achieves some great results
Efficient Streaming Language Models with Attention Sinks
Paper β’ 2309.17453 β’ Published β’ 14Note Check out this amazing blog post explaining this https://huggingface.co/blog/tomaarsen/attention-sinks
Stable Diffusion XL on TPUv5e
π2.04kGenerate images from text prompts
Note Run SDXL with TPU with a in-depth technical explanation
liuhaotian/llava-v1.5-7b
Image-Text-to-Text β’ Updated β’ 432k β’ 521Note A model that can do multimodal instruction following data
defog/sqlcoder2
Text Generation β’ Updated β’ 230 β’ 117Note Code models for the win! This is a 15B model that turns natural language to SQL
defog/sqlcoder-7b
Text Generation β’ Updated β’ 730 β’ 68Note And this is the 7B version of the above
-
MetaMath: Bootstrap Your Own Mathematical Questions for Large Language Models
Paper β’ 2309.12284 β’ Published β’ 18
meta-math/MetaMathQA
Viewer β’ Updated β’ 395k β’ 9.07k β’ 421Note A dataset of math questions for fine-tuning
AI Meme Generator
π₯111Create funny memes from images
Note Generate memes with IDEFICS, the multimodal model